Chapter 1: Introduction to the Eletrochemical and Photo-electrochemical Reduction of CO₂

The scale of the challenge in carbon dioxide conversion is enormous, with 35 gigaton of anthropogenic CO₂ generated by fossil fuel combustion/consumption every year. The average percentage of CO₂ in the atmosphere had risen to 403.38 ppm by September 2017 according to the data released by the National Oceanic and Atmospheric Administration of the US Department of Commerce,¹  with serious effects expected on global warming, climate patterns, sea levels, biodiversity, food production, human displacement, and so forth. There is a global consensus that immediate action must be taken to halt the relentless increase in atmospheric carbon dioxide levels, with the Global Apollo programme being a prime example of international initiatives being undertaken.²  In addition to the ever growing number of installations of renewable energy sources in different areas of the planet, there are great opportunities in “recycling”^3,4  carbon dioxide with biochars,⁵  by absorption,⁶  or by electrolysis powered by photovoltaic, wind, tidal, hydro-electric, salinity gradient energy, or the so-called “blue energy” systems.⁷  Although CO₂ sequestration has also been proposed,⁸  mimicking natural photosynthesis represents the most attractive but also scientifically challenging avenue.

In the context of artificial photosynthesis, links can be established between catalytic,⁹  electrochemical, and photo-electrochemical conversion of atmospheric CO₂, as schematically illustrated in Figure 1.1. Taking inspiration from nature, integrating functional units at the nanoscale capable of (i) capture solar light, (ii) CO₂ accumulation, and (iii) selective reduction to products can lead to new technologies that can be widely deployed for local mitigation of carbon emission. There are also integration approaches for “semi-artificial photosynthesis”, for example employing PS1 and PS2 apparatus extracted from cells and immobilised onto electrode surfaces. Although synergistic effects can be envisaged from such “nano-integrated” systems, such level of structural complexity currently rarely achieves high efficiency or stability. The electro-reduction of CO₂ is a tough problem so why combine this with the added complexity of an integrated solar cell? Why not just use conventional solar electricity without integration? So far, there are no technological developments that enable answering this question unambiguously. However, there is a large community of scientists, including those contributing to this book, that identifies a direct and selective path of CO₂ to a valuable carbon structure as one of the grand challenges in the path towards a low carbon economy.

This book primarily focuses on electrochemical conversion of CO₂, establishing correlation between the nature of the catalysts and the electrode potential. Figure 1.1 sketches how the potential bias (input energy) can be provided by solar energy either via photovoltaic solar cells,¹⁰  or by direct photo-generation of carriers at the electrocatalytic site.¹¹  These two approaches, illustrated by Figures 1.1A and B, respectively, represent two different levels of nano-integration. The third level (Figure 1.1C) involves the integration of CO₂ capture moieties directly to the catalysts active site in a combined unit. Depending on the catalyst properties, a variety of products can be generated directly (in situ), or by a separate process exploiting, for example, the formose reaction^12,13  to build up carbohydrates from formaldehyde. A large number of approaches has been presented in the literature for generating higher added value compounds based on various electrochemical reactors,¹⁴  homogeneous catalysts^15,16  or inorganic photosynthetic sites.¹⁷  Independently of the approach used for providing the driving force for CO₂ reduction, investigating the properties of the catalytic sites by (spectro)-electrochemical techniques provides extremely valuable mechanistic information. Some of the most recent developments in spectro-electrochemical tools are reviewed in Chapters 8–10.

Numerous reviews^18–20  and authoritative books^21–23  have been published on the topic of CO₂ reduction and electrocatalysis at electrode surfaces. The high symmetry of the carbon dioxide molecule has often been cited as a key contributor to the activation barrier, with a price in energy often paid by applying a high overpotential (excess potential with respect to the thermodynamic reduction potential). More recent studies, based on the so-called “scaling relations” formalism, point to the fact that the binding energy of intermediate species in multi-electron transfer reactions correlate in a linear fashion, resulting in large overpotentials for a variety of metals.^24–26  Azofra and Sun elaborate on this point in Chapter 6. At the fundamental level, a number of strategies reviewed in this book focus on breaking the challenge posed by this “scaling relation”.

From the thermodynamic point of view, the reduction of CO₂ to give useful hydrocarbons should occur under mild conditions, as summarised by eqn (1.1)–(1.5), and the corresponding Pourbaix diagrams²⁷  (see Figure 1.2). Thermodynamic data for organic media linking to aqueous media have also been reported.²⁸  Reduction to elemental carbon, and indeed nano-carbon products, is only accessible in molten salts and at high temperature.²⁹  The reversible potential for the CO₂/CO redox system (eqn (1.1)) has been elegantly confirmed by experiment (at pH 7) with enzyme-laden electrodes.³⁰ 

As mentioned previously, photoexcitation is the most attractive approach to generating highly energetic electrons capable of driving CO₂ reduction, as exemplified by the scheme in Figure 1.3, integrating light absorbing units (ruthenium bipyridyl photosensitiser), a metal oxide as electron accepting moiety, and carbon monoxide dehydrogenase.³¹ 

A key component in the overall process is the photo-absorber, which locally generates charge carriers. In addition to molecular dyes (e.g. ruthenium bipyridyl), other materials such as quantum dots, oxide nanoparticles,^32,33  carbon dots, and carbon based heterostructures.^34–36  With regards to catalytic centres, biological systems,³⁷  with metabolically engineered microorganisms,³⁸  and aided by synthetic biology, have been generating a tremendous amount of interest.^39,40  Other prime examples of nanoscale hybrid integrated systems involve immobilising “PS1 and PS2” centres extracted from cells,^41–43  or dehydrogenase enzymes⁴⁴  onto electrodes. Risbridger and Anderson review some of these strategies in Chapter 2 of this book.

Selectivity in the catalytic process can offer major advantages. Direct and selective catalytic transformations have been achieved (without additional catalyst) at semiconductor materials such as BiVO₄, for example, leading to high efficiency in the formation of (nearly) single products for methanol⁴⁵  (see Figure 1.4) or ethanol.⁴⁶  Also the formation of relatively complex molecules such as lactate,⁴⁷  or dimethylcarbonate in a single process has been reported.⁴⁸ 

Another dynamic area of research involves the development of electrochemical reactors for continuous flow CO₂ photo-reduction.⁴⁹  Conventional studies consist of catalytic electrodes placed in a reactor⁵⁰  in contact with a CO₂ saturated electrolyte solution, which could be either aqueous, organic, or ionic liquid media.⁵¹  The solubility of CO₂ in these systems is a crucial factor, particularly in aqueous media. The low CO₂ solubility in water can be somewhat alleviated under high pressure.⁵²  Centi et al. demonstrated an inverted or “driven” fuel cell reactor design with direct gas feed and a gas diffusion electrode, as illustrated in Figure 1.5, leading to non-Fischer–Tropsch product distribution on platinum catalysts.⁵³  Membrane reactor designs⁵⁴  with an effective triple phase boundary reaction zone can be employed to overcome solubility limitations.

Manipulating the activity of catalysts can be generally described in terms of (i) increasing the number of active sites, or (ii) the intrinsic activity of the material.⁵⁶  The first approach is based on the concept of nano-structuring,⁵⁷  as exemplified by the work of Broekman and co-workers employing the so-called Cu sponges.⁵⁸  However, entirely new catalysts and catalyst architectures are also needed. Intrinsic activity of late transition metals has been extensively investigated (see Figure 1.6) from the experimental^59–62  and theoretical points of view.

There are two distinct groups of catalysts with (i) CO forming metals (Cu, Au, Ag, Zn, Pd, Ga, Ni, and Pt) and (ii) formate forming metals (Pb, Hg, In, Sn, Cd, and Tl). Copper takes a special place with its capacity of generating a wide range of CO₂ reduction products and promoting carbon–carbon coupling. Although this classification relates to main generated products, there is a vast difference in the overpotentials required to drive the CO₂ reduction at these metals. Our understanding of the CO₂ reactivity at model surfaces has been significantly improved by detailed studies at single crystal electrodes, as reviewed in Chapter 4 of this book. New insights are possible employing sophisticated in situ spectroscopic techniques, as highlighted in Chapter 10. On the other hand, a lot less is understood about materials featuring highly correlated electrons, such as metal oxides.⁶⁴ 

The intrinsic activity of catalysts can also be modulated by alloying,⁶⁵  or by introducing metastable catalytic materials.⁶⁶  CO₂ reduction studies at ultrathin Pd layers grown on Au nanoparticles have provided interesting insights into the effect of the so-called electronic and strain effects.⁶⁷  Montes de Oca et al. have shown that decreasing the thickness of the Pd shells from 10 to 1 nm leads to an increase in the effective strain of the Pd lattice from less than 1 to 3.5%.⁶⁸  As illustrated in Figure 1.7, the reduction of CO₂ leads to the generation of mainly CO at the strain Pd shells, while relaxed shells produce HCOOH, CH₄, and C₂H₆.^67,69  In situ FTIR studies showed that the CO coverage and binding to Pd is strongly decreased at the thin Pd shells, which was rationalised in terms of an upward shift of the d-band centre due to lattice strain as estimated by DFT.⁷⁰  This strategy has been further extended with a large variety of core-shells and alloyed nanostructures.^71,72 

Transition-metal cations, including rare earth elements, play a key role in the activity of molecular catalysts.^73,74  The reactivity of these systems is also strongly affected by the ligand structure. For example, the catalytic reduction of CO₂ to CO at zinc porphyrinato complexes occurs through binding to the ligand rather than the metal centre.⁷⁵  Homogeneous catalysts based on coordination compounds will be discussed in Chapter 5.

Finally, the electrolyte salt (in particular the electrolyte cation) can be an important factor⁷⁶  as shown, for example, by the role of the alkali cation reported by Cuesta and coworkers.⁷⁷  Alkali cations have also been shown to affect the product distribution in CO₂ reduction.⁷⁸  Halide anions have been observed to improve the efficiency of CO₂ electro-reduction at copper surfaces.⁷⁹  Very intriguing are the effects of electrolyte cations in ionic liquid media, which strongly affect the structure of the double layer as well as enabling stabilisation of reaction intermediates such as CO₂^•− radical anions.⁸⁰  Zhao et al. reviews recent developments in CO₂ electro-reduction in ionic electrolytes in Chapter 7.

It is interesting to explore the transition zone from bulk to molecular catalyst materials. This is dominated currently by “nanoparticles”, which are known to be highly active catalysts, but also often poorly defined in terms of size distribution and surface chemistry. Single crystal nanoparticles are the exception. The increase in active surface area is crucial, but often also the electronic effects of going to sizes below 5 nm diameter can be important. The shape of nanoparticles is often crucial in catalysis^81,82  with novel “nano-frame” catalysts⁸³  providing further opportunities. In addition to changing the metal core, the ligand sphere can be tuned⁸⁴  to control catalytic performance. Microporous catalysts⁸⁵  and metal–organic framework catalysts⁸⁶  with high active surface area have been proposed.

At the level of metal cluster chemistry, there are also highly interesting candidates for catalysis with the added benefit of being molecularly well defined. Novel cluster systems, for example based on the Au₂₅ core, have been proposed^87–89  and shown to be highly active in the reduction of CO₂ to CO. The chemistry of heterogeneous and of homogeneous catalysis is now ever closer inter-linked with new nano-architectures being developed.

Confining homogeneous catalyst to interfaces offers new approaches combining the better understanding/tunability of molecular systems with the practicality of heterogeneous systems immobilised at electrode surfaces. Although methods based on conducting polymers and immobilised coordination polymers have been successfully developed in the past, there is a new emphasis on the architecture at nano-scale. A recent review by Reisner et al.⁹⁰  highlights the developments and promise in this field. Work by Domen and coworkers⁹¹  demonstrated the principle of combining a Ru(ii) dinuclear complex used for CO₂ reduction and a Ag-loaded TaON (Ag/TaON) semiconductor and light absorber. Active substrate materials such as Cu₂O are combined with surface immobilised Re metal complexes⁹²  to fundamentally change the pathway of the catalytic reaction (Figure 1.8). Similarly, the immobilisation of a molecular Mn catalyst on TiO₂⁹³  has been reported to allow stable photo-electro-reduction of CO₂ to CO in conjunction with a photo-anode generating oxygen.

A brief overview of current developments and the state-of-the-art in electrochemical reduction of carbon dioxide has been attempted. The existing breadth of catalysis materials and processes is impossible to fully cover and the development appears to be rapid. New computational theory and the new opportunity of computational artificial intelligence will open up further avenues for the sunlight-driven conversion of atmospheric CO₂ to useful fuels and products.

This book offers a broad and up-to-date perspective on topics including pure CO₂ reduction electrocatalysis, photo-electrocatalysis, the transition from homogeneous to heterogeneous catalyst systems, biological perspectives, in situ spectroscopy, and aspects of computational theory. There is a strong materials chemistry element in catalyst development and still a lot of adventure in catalyst discovery. The complexity of nano-integrated materials and technologies offer potential for economic rewards for the future, but also leads to frustratingly difficult research challenges based on finely controlled nano-architectures and well-understood reaction conditions.

Contributions to this book include in Chapter 2 “Bio-Inspired and Bio-Electrochemical Approaches in CO₂ Reduction Catalysis” by Thomas Risbridger and Ross Anderson with a focus on biological and bio-inspired processes for photo-electrochemical carbon dioxide reduction. This includes an overview of processes based in micro-organisms.

In Chapter 3 entitled “Copper Catalysts for the Electrochemical Reduction of Carbon Dioxide” by Hyung Mo Jeong, Boon Siang Yeo, and Youngkook Kwon copper is the focus. In this chapter reactivity trends and novel catalyst designs are critically assessed and an outlook on copper catalysis is provided.

Chapter 4 on “Single-Crystal Surfaces as Model Electrocatalysts for CO₂ Reduction” by Adam Kolodziej, Paramaconi Rodriguez, and Angel Cuesta introduces single crystal methods as a tool for better understanding of catalytic surface processes and for better linking experiments to theory and computational chemistry. This chapter focuses on platinum-group metals, copper, silver, and gold.

Chapter 5 entitled “Homogeneous M(bpy)(CO)₃× and Aromatic N-Heterocycle Catalysts for CO₂ Reduction” by Mitchell C. Groenenboom, Karthikeyan Saravanan, and John A. Keith offers an atomic scale perspective and view based on computational theory. Classic systems such as homogeneous metal complex catalysts and the group of N-heterocycle CO₂ reduction catalysts are investigated and explained.

In Chapter 6 “DFT Modelling Tools in CO₂ Conversion: Reaction Mechanisms Screening and Analysis” by Luis Miguel Azofra and Chenghua Sun the theory on heterogeneous catalyst system is added. Information is provided about “where”, “how”, and “why” in silico hypothesising and how to assess promising catalysts before the experimental testing of their catalytic performance.

Chapter 7 entitled “Electrocarboxylation in Ionic Liquids” by Shu-Feng Zhao, Michael D. Horne, Alan M. Bond, and Jie Zhang offers a perspective on CO₂ reduction in ionic liquid media. These media, often non-volatile and able to adsorb CO₂, offer a “game changer” in terms of room temperature electrocatalytic CO₂ reduction.

In Chapter 8 David E. Ryan and František Hartl introduce “IR Spectro-Electrochemistry and Group-6 α-diimine Catalysts of CO₂ Reduction” to demonstrate that new in situ reaction monitoring tools (in particular at low temperatures) can be highly powerful even for very complex homogeneous reaction pathways during carbon dioxide reduction reactions.

In Chapter 9 “Probing CO₂ Reduction Intermediates Employing in situ Spectroscopy and Spectrometry” by S. Pérez-Rodríguez, G. García, M. J. Lázaro, and Elena Pastor describe the state-of-the-art in monitoring reaction intermediates and catalytic pathways. A range of spectroscopy and spectrometry tools are reported and spectral properties of intermediates are linked to reactivity.

Finally, in Chapter 10 entitled “Surface-Selective and Time-Resolved Spectro-electrochemical Studies of CO₂ Reduction Mechanisms” by Alexander J. Cowan an overview is provided into surface-selective, surface-enhanced, and time-resolved forms of spectro-electrochemistry applied to mechanistic challenges in the field of CO₂ reduction.